SYSTEMS AND METHODS FOR ENHANCED DEPTH DETERMINATION USING PROJECTION SPOTS
Systems and methods for enhanced depth determination using projection spots. An example method includes obtaining images of a real-world object, the images being obtained from image sensors positioned about the real-world object, and the images depicting projection spots projected onto the real-world object via projectors positioned about the real-world object. A projection spot map is accessed, the projection spot map including information indicative of real-world locations of projection spots based locations of the projection spots in the obtained images. Location information is assigned to the projection spots based on the projection spot map. Generation of a three-dimensional representation of the real-world object is caused.
1 - 18 . (canceled)
19 . A method performed by a system of one or more computers, the method comprising:
obtaining, via an image sensor, images of a calibration board, the calibration board being positioned at a plurality of distances relative to a projector outputting a plurality of projection spots, the projection spots including a first projection spot;
identifying respective three-dimensional positions associated with each depiction of the first projection spot in the obtained images;
determining geometric information associated with a cone of light projected by the projector which forms the first projection spot in each of the plurality of images; and
generating a projection spot map which indicates, at least, information usable to determine a real-world location of the first projection spot as depicted in an image obtained by the image sensor.
20 . The method of claim 19 , wherein the projection spot map indicates information usable to determine real-world locations of the plurality of projection spots as depicted in images obtained by the image sensor.
21 . The method of claim 19 , wherein identifying a three-dimensional position associated with a depiction of the first projection spot comprises:
identifying a two-dimensional position associated with the first projection spot; and
based on camera parameters associated with the image sensor, identifying the three-dimensional position.
22 . The method of claim 19 , wherein geometric information for the cone of light includes a vector centered at an origin of the projector, the vector defining propagation of the cone of light.
23 . The method of claim 19 , wherein determining geometric information comprises:
fitting a line through a center of the three-dimensional positions associated with each depiction of the first projection spot, the line defining an axis of the cone of light.
24 . The method of claim 19 , further comprising:
after generating the projection spot map, obtaining a plurality of images of a real-world object, the images being obtained from a plurality of image sensors positioned about the real-world object, and the images depicting projection spots projected onto the real-world object via a plurality of projectors positioned about the real-world object;
accessing the projection spot map;
assigning location information to the projection spots based on the projection spot map; and
causing generation of a three-dimensional representation of the real-world object.
25 . The method of claim 24 , wherein the projection spot map includes information indicative of real-world locations of projection spots based on depictions of the projection spots in the obtained images, and wherein a depiction of a first projection spot comprises shape information, position information, and/or angle information, associated with image pixels which form the first projection spot.
26 . The method of claim 24 , wherein each projection spot represents an intersection of a cone of light projected by a projector with the real-world object, and wherein the projection spot map is based on geometric information associated with each cone of light.
27 . The method of claim 26 , wherein assigning location information to a first projection spot depicted in a first image comprises:
identifying first pixels of the first image which depict the first projection spot; and
identifying, based on the projection spot map and size, information indicative of a distance traveled by a cone of light associated with the first projection spot, and wherein the assigned location information is based on the indicated distance.
28 . The method of claim 24 , wherein assigning a location to a first projection spot depicted in a first image obtained from a first image sensor comprises:
identifying one or more pixels in the first image which form the first projection spot; and
identifying, for the first image sensor and the identified pixels, location information for the first projection spot based on the projection spot map.
29 . The method of claim 24 , wherein the assigned location information represents, at least, depth information for a surface of the real-world object, and wherein each projection spot depicted in an image represents an intersection of a respective cone of light from a respective projector with a portion of the surface of the real-world object.
30 . The method of claim 29 , wherein the assigned location information represents respective depths for each portion on which a respective cone of light is projected.
31 . The method of claim 30 , further comprising identifying an orientation associated with each portion, wherein identifying an orientation associated with a first portion on which a first cone of light is projected comprises:
identifying a shape associated with a first projection spot associated with the first cone of light, the first projection spot being depicted in one or more pixels of an image; and
identifying, based on the shape, the orientation of the first portion.
32 . A system comprising one or more processors and non-transitory computer storage media storing instructions that, when executed by the one or more processors, cause the processors to perform operations comprising:
obtaining, via an image sensor, images of a calibration board, the calibration board being positioned at a plurality of distances relative to a projector outputting a plurality of projection spots, the projection spots including a first projection spot;
identifying respective three-dimensional positions associated with each depiction of the first projection spot in the obtained images;
determining geometric information associated with a cone of light projected by the projector which forms the first projection spot in each of the plurality of images; and
generating a projection spot map which indicates, at least, information usable to determine a real-world location of the first projection spot as depicted in an image obtained by the image sensor.
33 . The system of claim 32 , wherein the operations further comprise:
after generating the projection spot map, obtaining a plurality of images of a real-world object, the images being obtained from a plurality of image sensors positioned about the real-world object, and the images depicting projection spots projected onto the real-world object via a plurality of projectors positioned about the real-world object;
accessing the projection spot map;
assigning location information to the projection spots based on the projection spot map; and
causing generation of a three-dimensional representation of the real-world object.
34 . The system of claim 32 , wherein the projection spot map indicates information usable to determine real-world locations of the plurality of projection spots as depicted in images obtained by the image sensor.
35 . The system of claim 32 , wherein identifying a three-dimensional position associated with a depiction of the first projection spot comprises:
identifying a two-dimensional position associated with the first projection spot; and
based on camera parameters associated with the image sensor, identifying the three-dimensional position.
36 . The system of claim 32 , wherein geometric information for the cone of light includes a vector centered at an origin of the projector, the vector defining propagation of the cone of light.
37 . The system of claim 32 , wherein determining geometric information comprises:
fitting a line through a center of the three-dimensional positions associated with each depiction of the first projection spot, the line defining an axis of the cone of light.